-
Controlling Executive Compensation Throughthe Tax Code
Gregg D. Polsky"
Table a/Contents
I. Introduction 878
II. Taxation of Executive Compensation 881A. Historical Tax
Treatment of Executive Compensation 881B. Enactment of § 162(m)
884
III. Two Models of Executive Compensation 886A. Agency Costs
887B. The Arm's Length Model 890C. The Managerial Power Model
891
IV. Effect of Section § 162(m) Under Each Model 892A. Arm's
Length Model 892
1. Avoidance 8932. Substitution Effect 896
a. Substituting Incentive Compensation for Cash 896b.
Substituting Formulaic Bonuses for
Discretionary Bonuses 8993. Deduction Forfeiture 9004. Director
Independence and Shareholder
Approval Requirements 9025. Effect on Nonaffected Firms 9026.
Conclusion Under the Arm's Length Model. 903
B. Managerial Power Model 9051. Avoidance 907
* Shiela M. McDevitt Professor of Law, Florida State University
College of Law. Ithank Kelli Alces, Barbara Banoff, Yariv Brauner,
Guy-Uriel Charles, Brian Galle, BrantHellwig, Claire Hill, Jeffrey
Kwall, Leandra Lederman, Martin McMahon, Jr., ChristopherPeterson,
Jim Repetti, Jim Rossi, Christopher Slobogin, David Walker, and
Ethan Yale for theirhelpful comments and suggestions on an earlier
draft.
877
-
878 64 WASH & LEE L. REV 877 (2007)
2. Substitution Effect 9083. Deduction Forfeiture 9124. Director
Independence and Shareholder
Approval Requirements 9135. Effect on Nonaffected Firms 9146.
Symbolic Value of§ 162(m) 9157. Conclusion Under the Managerial
Power Model 915
V. Empirical Studies 916A. Levels of Executive Compensation
917
1. Affected Firms 9172. Unaffected Firms 918
B. Substitution Effect 918C. Deduction Forfeiture 919D.
Conclusion 920
VI. Incidental Consequences 920A. Option Design 921B. Restricted
Stock 922C. Accounting Manipulation 923D. Inflation Effects 924E.
Conclusion 925
VII. Conclusion 925
1. Introduction
The topic of executive compensation has received a great deal of
recentattention from the news media, courts, and policymakers.
Newspapers havereported on the seemingly exorbitant pay packages of
chief executive officers(CEOs) of high profile companies.' Disney's
severance arrangement with itsformer CEO Michael Ovitz has made its
way through the Delaware court
1. See, e.g., Jenny Anderson, Goldman Chairman Gets a Bonus
of$53.4 Million, N.Y.TIMES, Dec. 20, 2006, at C2 (reporting the
$53.4 million year-end bonus paid by GoldmanSachs to Lloyd C.
Blankfein); Eric Dash, Compensation Experts Offer Ways to Help
CurbExecutive Salaries, N.Y. TIMES, Dec. 30,2006, at Cl (describing
the phenomenon of "runawayexecutive pay" and noting that CEO
compensation has increased more than 600% in the pasttwenty-five
years after proper adjustment for inflation); Mark A. Stein, Highly
Paid Chief isPaid $210 Million to Go Away, N.Y TIMES, Jan. 6,2007,
at C2 (discussing the $210 millionseverance package paid by Home
Depot to its former CEO Robert L. Nardelli).
-
CONTROLLING EXECUTIVE COMPENSATION 879
system.' The Securities and Exchange Commission recently
promulgatedlengthy new regulations aimed at improving the
transparency of certainexecutive pay components.' With the intent
to limit executive compensation,Congress is currently considering a
bill that would impose tax penalties ondeferred compensation
amounts in excess of $1,000,000.4 And PresidentGeorge W. Bush, in a
recent speech on Wall Street, exhorted boards ofdirectors to "step
up to their responsibilities" and ensure that "salaries andbonuses
ofCEOs [are] based on their success at improving their companies
andbringing value to their shareholders. ,,5
The topic of executive compensation has also generated
significantacademic discussion. For instance, in their highly
influential book Pay WithoutPerformance.' Professors Lucian
Bebchuck and Jesse Fried argue thatsignificant corporate governance
failures have led to unduly generous executivepay packages and
propose fundamental reform measures to resolve thesefailures.'
Their arguments and proposals stimulated a whole wealth ofacademic
literature. 8
Yet in all of this discussion, there has not been any
significant discussionor analysis of Congress's heretofore most
direct attempt to control executivecompensation amounts. In 1993,
during another period of intense criticism ofexecutive pay
packages, Congress enacted § 162(m) of the Internal RevenueCode,
which generally disallows public companies a deduction
fornonperformance-based compensation in excess of $1,000,000 paid
to the CEOand the next three highest paid executives. The purpose
ofthis legislation wasto enhance shareholder wealth in two ways: by
reducing the overall level ofexecutive compensation and by
influencing the composition of executivecompensation arrangements
in favor ofcomponents that were more sensitive tofirm
performance."
2. See In re The Walt Disney Co. Derivative Lit., 906 A.2d 27,
50 (Del. 2006)(concluding that the company's board of directors
complied with its fiduciary duties toshareholders in negotiating
and approving Ovitz's $130 million severance package).
3. 17 C.F.R. § 229.402 (2006).4. The Small Business and Work
Opportunity Act of2007, S. 349, 11Oth Congo (2007).5. Jim
Rutenberg, Bush Tells Wall St. to Rethink Pay Practices, N.Y.
TIMES, Feb. 1,
2007, at Cll.6. LUCIAN BEBCHUK & JESSE FRIED, PAY WITHOUT
PERFORMANCE: THE UNFULFILLED
PROMISE OF EXECUTIVE COIv1PENSATION (2004).7. See id. at 1-12
(providing an overview of corporate governance concerns).8. See
William W. Bratton, The Academic Tournament over Executive
Compensation, 93
CAL. L. REv. 1557, 1559 (2005) (describing the "academic
tournament" stimulatedby Bebchuk& Fried).
9. See Robert M. Halperin, Young K. Kwon & Shelley C.
Rhoades-Catanach, The
-
880 64 WASH. & LEE L. REV 877 (2007)
During this time ofmounting criticism ofexecutive compensation
norms,there is political pressure on Congress to take further
measures to intervene inthe relationship between the corporation
and its managers. One such measureis the $1,000,000 limit on
deferred compensation described above." Othersimilar proposals are
pending. 11 Before taking these new measures, however,Congress
ought to first reflect upon the effectiveness of § 162(m) in
achievingits intended results.
The primary goal ofthis Article is to evaluate the efficacy of §
162(m). Iconsider the likely effect of § 162(m) under the two
currently prevailing (butopposing) views ofhow executive
compensation arrangements are negotiated inthe public company
context.V Under one view (the arm's length model), thecorporation's
board ofdirectors negotiates effectively on behalfofshareholdersin
setting management compensation. 13 Under the other view (the
managerialpower model), as a result of structural biases and other
problems, the board iscaptured by management and is therefore
willing to overpay management. 14Under the managerial power model,
the only meaningful constraint on directorsand management is the
risk that relevant outsiders would perceive thecompensation amounts
as egregious." To mute this constraint, the board and
Impact ofDeductibility Limits on Compensation Contracts: A
Theoretical Examination, 23 J.AM. TAX. ASS'N 52, 52 (Supp. 2001)
(noting these objectives of § 162(m)); John Graham &Yonghan
Julia Wu, Executive Compensation, Interlocked Compensation
Committees and the162(m) Cap on Tax Deductibility 7 (Jan. 2007)
(unpublished working paper, on file with theWashington and Lee Law
Review), available at http://ssm.com/abstract=959686
(liThelegislation was enacted in response to the perceived
excessive top executive pay, and intendedto promote broader
application of performance-based compensation ....").
10. See supra note 4 and accompanying text (describing a
Congressional proposal toimpose tax penalties on deferred
compensation amounts in excess of $1,000,000).
11. See, e.g., H.R. 1257, 110th Congo (2007) (requiring that
shareholders be allowed tovote on executive compensation
arrangements); H.R. 3260, 110th Congo (2007) (denyingcompany
deductions for executive compensation in excess of twenty-five
multiplied by thesalary of the company's lowest paid worker); S. 2,
11OthCongo (2007) (expanding the scope ofemployees whose pay is
subject to § 162(m)).
12. This Article does not specifically address whether the twin
goals of § 162(m) (i.e.,reducing the level of executive pay and
increasing the performance-sensitivity of such pay) areworthwhile
legislative pursuits. As discussed in more detail below, under the
arm's lengthmodel, § 162(m) is unnecessary to achieve optimal
compensation arrangements. Under themanagerial power model, pay is
generally too high and too performance-insensitive;accordingly,
under that model the goals of § 162(m) are beneficial. The question
under themanagerial power model is whether the goals of § 162(m)
can realistically be achieved in lightof how the provision actually
operates.
13. See BEBCHUK & FRIED, supra note 6, at 17-20 (describing
the arm's length model).14. See id. at 61-80 (describing the
managerial power model).
15. See id. at 65 ("Whether directors and managers will be
deterred from adopting a givencompensation arrangement depends on
the extent to which it will be viewed by relevant
-
CONTROLLING EXECUTIVE COMPENSATION 881
management collude to disguise and understate the true value of
executivecompensation, according to adherents of the managerial
power model. 16
Ultimately, this Article concludes that, under either model, §
162(m) islikely ineffective. Instead of increasing shareholder
wealth by limitingexecutive compensation and aligning executive
incentives, § 162(m) probablyhas the opposite effect ofdecreasing
shareholder wealth. Under the managerialpower model, this perverse
result is possibly even more pronounced.
In addition to predicting the likely effect of § 162(m), this
Articleconsiders empirical studies of its actual impact since its
enactment almostfifteen years ago. Most of the evidence is
consistent with predictions underboth models, while other evidence
is more consistent with one or the othermodel under certain
conditions.
Finally, the Article describes some unintended incidental
effects of theprovision. Section 162(m) discourages the use of
certain compensationstructures that are arguably more efficient,
while favoring other flawedstructures. It encourages the use of
formulaic bonus arrangements, whichthemselves encourage accounting
manipulation. Furthermore, the scope of§ 162(m) is expanding
through legislative inaction because the provision is notadjusted
for inflation.
This Article proceeds as follows: Part II describes the
historical treatmentof executive compensation under the Code and
then discusses the backgroundbehind, and the operation of, §
162(m). Part III discusses the agency problemsinherent in the
executive/firm relationship and the two competing models ofhow
executive compensation arrangements are negotiated. Part IV
predicts thelikely effects of § 162(m) under each of the competing
models. Part V thenconsiders the results ofempirical studies of §
162(m). Part VI describes someunintended incidental consequences of
the provision. Part VII concludes.
II. Taxation ofExecutive Compensation
A. Historical Tax Treatment ofExecutive Compensation
The Code allows a deduction only for "reasonable" compensation
paid toemployees." While the text of the statute suggests that all
compensation paidto an employee must be evaluated for its
reasonableness, in practice this is notthe case. In fact, the
reasonableness factor is almost exclusively used by the
outsiders as unjustified or even abusive or egregious. ").
16. Id. at 67-70.17. I.R.C. § 162(a) (2000).
-
882 64 WASH. & LEE L. REV 877 (2007)
IRS to deny deductions for payments that are in substance gifts
or dividends butare disguised as compensation."
For example, if a doctor subject to a 400/0 marginal tax rate
wanted tomake a $10,000 gift to her college-aged son, the doctor
could overpay the son$10,000 for work performed for her medical
practice and attempt to deduct theamount. If this strategy were
successful, the $10,000 excess "salary" wouldsave the doctor $4,000
in taxes. 19 If the son were in a taxable position.i"
theoverpayment would increase the son's taxes;" however, so long as
the son'smarginal tax rate was less than 40%, the family unit's
aggregate tax liability isreduced at the expense of the fisc.22 The
IRS would attack this strategy byarguing that the son's
compensation was unreasonably high, which, ifsuccessful, would
result in the denial of the doctor's deduction.
A similar tax avoidance strategy is sometimes used by closely
heldcorporations where the principal shareholders are also the
principal employees.Ifthe shareholders/employees take cash out
ofthe company through payment ofdividends, the corporation gets no
deduction for the dividends." On the otherhand, if the
shareholders/employees take cash out of the company throughpayment
ofsalaries and bonuses, the corporation is allowed a deduction for
thecompensation payments (provided that the corporation's
characterizations ofthe payments are respecredj" Recognizing this,
the IRS attempts to protect the
18. See 1 BORIS I. BITTKER & LAWRENCE LOKKEN, FEDERAL
TAXATION OF INCOME,ESTATES, AND GIFTS ~ 22.2.1 (3d ed. 1999)
(noting that the IRS "almost never" uses thereasonableness
condition to deny deductions for salary unless the risk of
disguised gift ordividend is present).
19. The deduction would absorb $10,000 of otherwise taxed
income, which at the 40%marginal tax rate would have resulted in an
increased tax liability of $4,000. To keep theanalysis ofthis
hypothetical simple, I ignore the impact offederal employment
taxes, which can,in certain circumstances, be a significant
consideration.
20. The son might not be in a taxable position because he might
not have income thatexceeds the sum ofhis standard deduction,
personal exemption, and other deductions (e.g., thededuction for
higher education expenses in § 222). In addition, the son may have
credits (suchas the HOPE and Lifetime Learning credits for
education expenses) that could absorb anyremaining tax
liability.
21. If the doctor had simply gifted the $10,000, the son would
not pay tax on the amountbecause I.R.C. § 102 excludes gifts from
gross income.
22. For example, assume that the son is in the 15% marginal tax
bracket. If the doctorsimply gifted the $10,000, the doctor would
pay $4,000 of tax on the $10,000 because shereceives no deduction.
The son would not owe any tax because he has received a gift. If
theoverpayment strategy is successful, the doctor would owe no tax
on the $10,000 (because shereceives a deduction which offsets the
income) while the son would owe $1,500 in additionaltax (i.e.,
$10,000 x 15%). If the overpayment strategy is successful, the
family unit saves$2,500 in taxes (Le., $4,000 $1,500).
23. I.R.C. § 311(a) (2000).24. The benefits of bailing out
corporate earnings through payments characterized as
-
CONTROLLING EXECUTIVE COMPENSATION 883
tax base by ensuring that purported salary payments to
shareholder/employeesare not in substance disguised dividends,
relying on the reasonableness factor todo the heavy lifting in this
regard. A critical factor in the reasonablenessanalysis in the
corporate context is therefore the degree ofoverlap between
theemployees and the shareholders." Ifthe corporation has four 25%
shareholderswho each perform similar services for the corporation,
the risk of disguiseddividends is very high. If, on the other hand,
only one ofthe four shareholdersis an employee of the corporation,
the risk of disguised dividends is very lowbecause the nonemployee
shareholders would generally be unwilling to allowthe one
employee/shareholder to siphon offtheir share ofsome ofthe
corporateearnings for his exclusive benefit. 26
In the public company context, concerns about disguised gifts or
dividendsare simply not present, and therefore there is no need for
the IRS to scrutinizethe level ofexecutive compensation to protect
the tax base. One could certainlyargue that executive compensation
is not set by an arm's length process and,therefore, that part of
executive pay constitutes "rents" extracted bymanagement as a
result of this flawed process." However, even these rentswould
generally constitute an ordinary and necessary expense that should
bedeductible by the corporation in arriving at corporate taxable
income." Theserents are simply a cost ofdoing business as a public
company where ownershipand control are separate by virtue
ofadministrative necessity." Conceptually,therefore, compensation
paid to executives of public companies ought to bedeductible in
full, even if part of the compensation can properly becharacterized
as renta'" This is consistent with the general practice ofthe
IRS
compensation instead of dividends have been reduced by the tax
rate cut on dividends enactedin 2003, see I.R.C. § l(h)(II), which
is scheduled to expire on December 31,2010. Pub. L. No.109-222, §
102, 119 Stat. 365 (2003). For analysis of this impact of the
dividend tax rate cut,see Alan L. Feld, Dividends Reconsidered, 101
TAXNOTES 1117 (2003).
25. See PAUL R. McDANIEL, MARTIN 1.MCMAHON, JR. & DANIEL L.
SIMMONS, FEDERALINCOME TAXATION OF CORPORATIONS 228 (3d ed. 2006)
(noting the importance ofproportionality in this context).
26. This assumes that the four shareholders are unrelated. If
the shareholders are related,then there is a risk that the
transaction is a disguised dividend combined with a disguised
gift(from the donor/shareholder who would be entitled to the
dividend to the donee/employee whoreceives the "compensation").
27. See generally BEBCHUK & FRIED, supra note 6 (discussing
the corporate governanceconcerns implicated by executive
compensation arrangements).
28. See I.R.C. § 162(a).29. Alternatively, one could
characterize the rent extraction as theft by the executives.
Since theft losses are deductible, Treas. Reg. § 1.165-1 (d)(3)
(1996), the payment ofthese rentsby the corporation would still be
deductible under this view.
30. Other considerations support this conclusion. It would be
administratively quite
-
884 64 WASH. & LEE L. REV 877 (2007)
to scrutinize compensation arrangements for corporate executives
only in theclosely held context where there is a significant
overlap between thoseindividuals who work for the company and those
individuals who own shares inthe company."
B. Enactment of§ 162(m)
In 1993, in response to public outcry over the level and
performance-insensitivity ofexecutive pay, Congress added § 162(m)
to the Code." Whilethis provision was sometimes justified by its
proponents as denying a "subsidy"for excessive executive pay, most
proponents properly characterized it as apenalty for what they
considered the illegitimate compensation practices ofpublic
companies.r' Section 162(m) was intended both to reduce the
overalllevel ofexecutive compensation and to make such compensation
more sensitiveto firm performance.34
It is clear that § 162(m) is not grounded in tax policy
considerations; asnoted above, executive compensation ought to be
deductible in full regardlessof whether some portion constitutes
rents." Rather, the provision is simply apenalty that is
administered through the tax code."
difficult to distinguish between the compensation versus rent
portions of an executive'sremuneration. In addition, depending on
the incidence of the corporate tax, denying thededuction for the
rent portion might only add to the agency cost borne by
shareholders: Theshareholders first overpay for executive services
and then they may incur the brunt of the extracorporate tax
attributable to denying the deduction for such overpayment.
31. See 1 BORIS 1. BITTKER & LAWRENCE LOKKEN, FEDERAL
TAXATION OF INCOME,ESTATES, AND GIFTS ~ 22.2.2 (3d ed. 1999)
(noting that "[v]irtually all reasonable compensationcases involve
family corporations or other closely held enterprises").
32. See Ryan Miske, Note, Can't Cap Corporate Greed: Unintended
Consequences ofTrying to Control Executive Compensation Through the
Tax Code, 88 MINN. L. REv. 1673,1686 (2004) (noting that § 162(m)
was enacted "after the populist outrage over executivecompensation
reached a high during the 1992 presidential race").
33. Allowing a corporate deduction for executive compensation,
even where part of thepay consists of rents, is not a subsidy. As
noted above, even the rent portion of thecompensation ought to be
deductible in arriving at corporate taxable income because the
rentsare a cost of doing business. See supra notes 29-30 and
accompanying text. As a result,disallowing the deduction is not
properly characterized as the withdrawal ofa subsidy;
instead,disallowing the deduction is best viewed as the imposition
of a penalty.
34. Supra note 9 and accompanying text.
35. Supra notes 29-30 and accompanying text.
36. Section 162(m) and similar provisions (e.g., § 280G (denying
deductions forexcessive golden parachute payments) and § 162(1)
(denying deductions for fines paid» are thusanalogous to the whole
host of tax expenditures, provisions in the Code which are designed
tosubsidize favored activities. The classic example of a tax
expenditure might be the home
-
CONTROLLING EXECUTIVE COMPENSATION 885
In general, § 162(m) disallows a deduction for certain
compensation("nonqualified" compensation) in excess of $1,000,000
paid by publiccompanies to its CEO and next three highest paid
executives. The $1,000,000limit is applied annually." For purposes
of § 162(m), qualified compensationis not subject to the $1,000,000
limit.38 Qualified compensation isremuneration that is paid
pursuant to a plan that (i) provides for payments basedon objective
performance goals, (ii) is approved by an independentcompensation
committee ofthe board, and (iii) is approved by a majority
oftheshareholders after disclosure of the material terms of the
compensationarrangement." With respect to stock options and stock
appreciation rights,"they are generally considered qualified
compensation if they are not "in-the-money?" at the time of grant,
provided that the independent director and
mortgage interest deduction, which allows a deduction for an
otherwise nondeductible expensein order to subsidize the activity
of borrowing to purchase owner-occupied housing. "Negative"tax
expenditures, like § 162(m), are designed to penalize disfavored
behavior.
37. I.R.C. § 162(m)(I) (2000). Technically, the statute applies
to compensation earned bythe CEO and any other employee whose
compensation "is required to be reported toshareholders under the
Securities Exchange Act of 1934 by reason of such employee
beingamong the 4 highest compensated officers for the taxable year
(other than the chief executiveofficer"). I.R.C. § 162(m)(3)
(emphasis added). Because the SEC's recently issued rulesrelating
to executive compensation now require the reporting of compensation
earned by theCEO, the CFO, and the next three highest compensated
executives, the IRS has indicated that itwill interpret § 162(m)(3)
to apply only to compensation earned by the CEO and the next
threehighest compensated executives. See U.S. DEP'TOF THE TREASURY,
INTERNAL REVENUE SERVICEBULLETIN, NOTICE 2007-49 (2007), available
at http://www.irs.gov/irb/2007-25_irb/arI2.html("The IRS will
interpret the term 'covered employee' for purposes of § 162(m) to
mean anyemployee [whose total compensation] is required to be
reported to shareholders under theExchange Act by reason of such
employee being among the 3 highest compensatedofficers ....").
38. I.R.C. § 162(m)(4)(C).
39. Id.40. A stock option is the right to purchase a share of
stock at a specified "exercise price"
(usually the fair market value on the date ofgrant) during a
specified period oftime (usually tenyears from grant). A stock
appreciation right (SAR) gives the holder the right to a
paymentmeasured by the difference between the fair market value of
the stock on the date of exerciseand a specified price (usually the
fair market value of the stock on the date of grant). SARsusually
have a term often years. Because employees usually sell the stock
received immediatelyupon option exercise, stock options and SARs
are functionally equivalent from the standpoint ofthe employee.
From the employer's perspective, SARs (unlike stock options) do not
dilute thenumber of shares outstanding. Both stock options and SARs
usually vest over time, meaningthat they become exercisable only
after a specified period of employment with the firm.
41. An option is not in-the-money if the exercise price is equal
to or greater than the fairmarket value of the underlying stock at
the time of grant.
-
886 64 WASH. & LEE L. REV 877 (2007)
shareholder approval requirements are met with respect to the
plan under whichthe options or rights are granted.42
Prior to the enactment of § 162(m), executive bonuses were
generally paidat the discretion of the board of directors. In order
for bonus arrangements toqualify under § 162(m), they must be
objectively determinable." However, theregulations provide that
boards may retain "negative discretion"-Le., theability to reduce
bonus amounts that are derived from objective formulas-without
impairing the qualified status of the bonus plan."
Under current compensation practices, the most common types
ofnonqualified compensation that trigger § 162(m) are salaries and
discretionarybonuses." Amounts paid under objective bonus plans
(either with or withoutnegative discretion ofthe board) and stock
option grants are the most commontypes of qualified
compensation."
III. Two Models ofExecutive Compensation
This Part will describe the two competing (though not
whollycontradictory) views of how executive compensation is
negotiated." Under
42. With respect to the stock option plan, shareholders need
only approve the totalnumber of options subject to the plan, the
maximum number of options that an employee mayreceive under the
plan, and the terms of the options.
43. See Treas. Reg. § 1.162-27(e)(2)(ii) (1996) ("A formula or
standard is objective if athird party having knowledge ofthe
relevant performance results could calculate the amount tobe paid
to the employee. In addition, a formula or standard must specify
the individualemployees or class of employees to which it applies.
").
44. Treas. Reg. § 1.162-27(e)(2)(iii)(A).
45. See Nancy L. Rose & Catherine Wolfram, Regulating
Executive Pay: Using the TaxCode to Influence ChiefExecutive
Officer Compensation, 20 J. LAB. ECON. 138, 141 (2002)(discussing
the effects of § 162(m) on executive compensation).
46. See id. (discussing qualified performance-based
compensation).
47. See Bebchuk and Fried, in an article co-authored with David
Walker, describe therelationship between the two models as
follows:
Although the managerial power approach is conceptually quite
different from the[arm's length] approach, the former is not
proposed as a complete replacement forthe latter. One can take the
view that compensation arrangements are shaped bothby managerial
power and by what would be optimal. The managerial powerapproach
merely implies that compensation practices cannot be
adequatelyexplained by optimal contracting alone. Rather, practices
might be adopted thatdeviate significantly from those suggested by
optimal contracting. Under themanagerial power approach,
compensation practices can be fully understood onlywith careful
attention to the role of managerial power.
Lucian A. Bebchuk, Jesse M. Fried & David I. Walker,
Managerial Power and Rent Extractionin the Design ofExecutive
Compensation, 69 U. CHI. L. REv. 751,755 (2002).
-
CONTROLLING EXECUTIVE COMPENSATION 887
one view, executives negotiate their pay with a board
ofdirectors that strives toget the best deal possible for
shareholders, thereby maximizing firm value."Under the other,
executives negotiate with a compliant board ofdirectors that
iswilling to overpay management as long as it will not raise red
flags to relevantoutsiders." Before describing these two models in
more detail, the agencyproblems inherent in the executive-firm
relationship are discussed.
A. Agency Costs
In any employment relationship, there exist conflicts of
interest betweenthe parties. The most significant is usually the
problem ofshirking. 50 All else(e.g., pay) being equal, employees
would generally desire to give less effort, tothe detriment of the
employer. If this problem is significant, the employercould (ex
ante) negotiate for some or all of the employee's pay to be based
onsome performance metric, thus increasing the alignment of the
employer andemployee's interests." The employee would presumably
require a higheramount oftotal compensation with this new
compensation mix because: (i) theadjusted compensation mix no
longer allows the employee to impose the fullcost of shirking on
the employer and (ii) the adjustment to the compensationarrangement
will often shift risk from the employer to the employee, who
isgenerally more risk-averse than the employer.52
48. See BEBCHUK & FRIED, supra note 6, at 17-20 (describing
the arm's length model).Negotiating the best deal for shareholders
does not necessarily mean obtaining the executive'sservices for the
cheapest price. It may mean negotiating compensation structures
that help toalign the executive's interests with the shareholders'
interests. See id. at 19 (suggesting one wayto align such interests
is to base the CEO's compensation in part on increasing
shareholdervalue).
49. See id. at 61-79 (describing the managerial power
model).
50. See Stephen M. Bainbridge, Executive Compensation: Who
Decides?, 83 TEX. L.REv. 1615, 1620-23 (2005) (describing the
problem of shirking by employees).
51. See id. (discussing solutions for the problem of
shirking).
52. See id. If the employee could perfectly and costlessly hedge
this additional risk by,for example, purchasing a financial
derivative, then this factor would be negated. In the publiccompany
executive context, there are significant contractual, securities
law, and tax lawconstraints that either prohibit hedging strategies
or make them unduly costly to implement. SeeDavid M. Schizer,
Executives and Hedging: The Fragile Legal Foundation of
IncentiveCompatibility, 100 COLUM. L. REv. 440,459-93 (2000)
(describing the existing contractual,securities law, and tax
barriers to hedging); see also Brian J. Hall & Kevin J. Murphy,
TheTrouble with Stock Options, 17 J. ECON. PERSP., 49, 55 (2003)
("Press accounts sometimesclaim that executives can effectively
hedge or unwind the risk oftheir options through
financialtransactions (involving derivatives) with investment
banks. But we know of no evidenceshowing that this practice is
widespread, and our many conversations with
knowledgeablepractitioners suggest that this practice is quite
rare. "). As a result, this Article ignores the
-
888 64 WASH. & LEE L. REV. 877 (2007)
For example, assume that MGM is negotiating with Tom Cruise to
starin an upcoming movie and that MGM would be willing to pay
Cruise$20,000,000 cash. Also assume that Cruise would be willing to
accept$20,000,000 cash. However, MGM is concerned that, if it were
to pay onlycash, Cruise would give less than optimal effort in
making and promotingthe movie. IfMGM instead offers Cruise
nontransferable royalty rights inthe movie that have an objective
ex ante fair market value of $20,000,000,he would likely value
those rights as worth less than $20,000,000 cashbecause he could no
longer shirk at the sole expense ofMGM and becausethe cash is for
the most part risk-free.t' while the royalty rights mayultimately
turn out to be worth much less than $20,000,000. 54 By makingCruise
effectively a partner in producing the movie, risk is shifted
fromMGM (a more efficient risk-bearer) to Cruise (a less efficient
risk-bearer). 55As a result, we might expect to see the parties
strike a bargain where someamount ofcompensation is cash and some
is performance-based and wherethe total expected value of the
compensation exceeds $20,000,000. Suchan arrangement would maximize
the expected value to the parties aftertaking into account the
possibility of shirking and their respectivepreferences regarding
risk.
In an arm's length arrangement (such as the MGM/Cruise
contract),we would expect that the parties would reach the optimal
mix ofperformance and nonperformance based compensation. As a
result,lawmakers do not have much interest in regulating the
MGM/Cruisecompensation arrangement, a deal struck between two
sophisticated parties,
possibility of executive hedging.
53. Of course, if the cash payment is deferred until after
Cruise performs, Cruise is subjectto the risk that MGM will become
bankrupt or insolvent before he is paid. Even if Cruise ispaid
simultaneously with or before performance, Cruise will risk
incurring opportunity costs ifMGM goes bankrupt between the time he
agrees to do the movie and the time he realizes thatMGM will not
fully perform under the agreement (because Cruise could not agree
to appear inany other movies during the time he is required to work
on MGM's). IfMGM is quite solvent,these risks are
insignificant.
54. The royalty rights could also turn out to be worth much more
than $20,000,000.
55. MGM is more efficient at bearing risk for a variety of
reasons. First, MGM makesmore movies than Cruise; therefore, its
investment in the particular movie is more diversifiedthan
Cruise's. Second, MGM's owners (i.e., the owners of Disney) are
themselves able todiversify by making investments in other
companies. Cruise, meanwhile, has a relatively large,undiversified
investment in the movie. He will devote substantial time and effort
in making themovie, and the movie may have a significant impact on
his reputation and future income-producing capacity.
-
CONTROLLING EXECUTIVE COMPENSATION 889
even though some might think that Cruise is grossly overpaid. 56
Policymakers are content to let the market work. 57
In the senior executive/firm context, another significant agency
probleminvolves the willingness ofcorporate policy-makers to cause
the firm to take onadditional risk." Managers are typically more
risk-averse than shareholders ofthe firm with respect to
firm-specific risk because managers make anundiversified investment
of human capital in the firm. If the firm does verypoorly, managers
will lose their job and status, and their reputation and
futureincome-producing capacity could be adversely affected.
Meanwhile,shareholders can readily diversify their firm-specific
risk by investing in otherfirms; they are therefore much less
risk-averse than managers." As a result,managers might choose
projects that entail a level of risk that is sub-optimalfrom the
shareholders' perspective. In addition, managers might choose
tofinance projects using equity or existing cash instead ofdebt
even though debt-financing (which is more risky than
equity-financingj'" might best increase firmvalue.
Like the problem of shirking, risk-based agency costs can, in
theory, beameliorated by substituting performance-based
compensation for cash. If anexecutive is paid only cash, she will
get the amount of cash she is due, nomatter how well the firm
performs, so long as it does not become insolvent. Inthis regard,
the executive is akin to an unsecured creditor, which is subject
onlyto insolvency risk but does not participate in corporate growth
to any extent.By giving the executive performance-based
compensation, the executive nowdoes participate in corporate
growth." As a result, the executive is more likelyto promote a
corporate policy of optimal risk-taking.
56. Ofcourse, policymakers might be concerned about
redistributing wealth. Historically,such redistributive efforts in
the United States have focused on relative amounts of wealth
(orincome) and not on their source (e.g., entrepreneurial efforts,
investment returns, inheritance,luck, or a favorable employment
contract).
57. In the senior executive/firm context, it may be questioned
whether shirking in fact is asignificant problem. Do CEOs respond
to performance-sensitive pay with increased effort?And, if so, do
increases in CEO effort at the margin translate into increased firm
performanceafter taking into account the additional compensation
required by the CEO to offset theadditional risk incurred by her in
accepting performance-sensitive pay in lieu of cash? Fordiscussion
of these issues, see Bainbridge, supra note 50, at 1632-35.
58. See id. at 1621.59. See Melvin A. Eisenberg & Brett H.
McDonnell, Expectation Damages and the
Theory ofOverreliance, 54 HASTINGS LJ. 1335, 1366 n.36
(2003).60. Debt-financing increases the leverage of the firm, which
increases the reward to
shareholders if the company is profitable enough to service the
debt. On the flip side, increasingleverage increases the risk of
firm insolvency or bankruptcy.
61. As noted above, this Article ignores the prospect
ofexecutive hedging strategies due
-
890 64 WASH & LEE L. REV. 877 (2007)
Certain performance-based structures are better than others in
amelioratingthe risk-based agency costs. For example, options are
quite beneficial in thisregard. As noted above, executives are more
risk-averse than shareholdersbecause of their large and
undiversified human capital investment in the firm.Option-holders,
on the other hand, are more risk-preferring because, while
theyparticipate in share increases above the exercise price, they
do not suffer fromshare decreases below the exercise price (no
matter how great).62 As a result,compensatory options should, in
theory, counteract the effect ofexecutive risk-averseness. On the
other hand, formulaic-based bonus arrangements may notbe as useful
in ameliorating the risk-based agency costs. Researchers havefound
that these bonus arrangements typically "result in a greater
penalty forextreme negative performance and a smaller reward for
extreme positiveperformance. ,,63 Thus, these arrangements will not
be as helpful incounteracting executive risk-averseness.
In summary, while shirking and risk-based agency costs push at
themargin in favor of greater performance-based compensation
relative to cashcompensation, the executive's risk-averseness
pushes in the other direction. Assuch, determining the optimal
level of performance-based compensationrelative to salary is a
difficult task.
B. The Arm's Length Model
Under the arm's length model, the board ofdirectors negotiates,
on behalfof shareholders, just as effectively as the parties in the
hypotheticalMGM/Cruise arrangement. Under this model, the market
will take into accountthe competing factors, ultimately resulting
in the optimal overall level ofexecutive compensation as well as
the optimal mix of performance-basedcompensation and salary."
to legal constraints that make them difficult or impossible to
implement. See supra note 52.62. Option holders are more
risk-preferring than stockholders because, while option
holders participate in corporate growth above the exercise price
just like shareholders, they(unlike shareholders) do not suffer
from reductions in stock price below the exercise price.Thus,
compensatory options may balance out to some extent the impact of
executive risk-averseness.
63. Austin Reitenga et al., CEO Bonus Pay, Tax Policy, and
Earnings Management, 24 J.AM. TAX. Ass 'N 1, 2 (Supp. 2002). This
results from the fact that most objective bonusarrangements use a
formula whereby the bonus is based on results within 80% to 120% of
thetargeted performance. As a result, performance above 120% does
not increase the bonuspayment, while performance below 80% results
in no bonus.
64. In light of the cozy relationship between boards and
management, the arm's lengthmodel might appear naive. However, it
must be remembered that the market should in theory
-
CONTROLLING EXECUTIVE COMPENSATION
C. The Managerial Power Model
891
On the other side ofthe spectrum is the managerial power model,
which isbased on the assumption that primarily because of
structural bias, directors donot bargain at arm's length with
managers." As a result, the model suggeststhat executive
compensation structures are more generous and
lessperformance-sensitive than those that would be the product of
arm's lengthbargaining. The structural bias results primarily from
the fact that directors areoften, as a de facto matter, appointed
by management." In addition, becausemanagement and directors work
closely together and often run in the samesocial circles, it is
argued that directors do not negotiate as hard as they wouldwith
truly adverse parties." Finally, because directors are
part-timeindependent contractors, they are unwilling to devote the
time necessary toeffectively negotiate and design executive
compensation structures, accordingto adherents of the managerial
power model."
Under the managerial power model, the only significant
constraint toexcessive management compensation is the risk of
outrage "shared by thoseoutsiders whose views matter most to [the
directors and cxecutivesj.t''"Adherents of the managerial power
view argue that, to ameliorate this risk,directors and management
devote considerable energy in devising ways tocamouflage the real
value of executive compensation arrangements."
react adversely to poorly structured or overgenerous executive
compensation structures. As aresult, all else being equal, capital
should flow into firms with "good" compensationarrangements and
away from ones with "bad" compensation arrangements, resulting
incorresponding changes to market price of firms. Such a market
reaction would act as a check inkeeping the parties honest in
negotiating executive pay. Whether this in fact occurs has been
thesubject ofvigorous debate. Compare Bainbridge, supra note 50, at
1635-37 (arguing that theabsence of contractual constraints on
managerial power in the IPQ context "is an
important-anunanswered-strike against the managerial power model"),
with BEBCHUK & FRIED, supra note6, at 53-58 (arguing that while
the market does exert influence, this influence does not
preventresults that depart from those obtained through arm's length
bargaining).
65. See BEBCHUK & FRIED, supra note 6, at 61-65 (describing
the managerial powermodel).
66. See id. at 25 (arguing that while a desire to be reelected
should encourage directors tobe attentive to shareholder needs, the
reality is that the director slate proposed by companymanagers is
almost never challenged).
67. See id. at 31-34.68. See id. at 36-37.69. See id. at 66.70.
See id. at 67 ("The main costs to directors and managers of
adopting compensation
arrangements that favor managers, then, depend mainly not on how
costly the arrangementsactually are to shareholders, but on how
costly the arrangements are perceived to be byimportant outsiders.
").
-
892 64 WASH & LEE L. REV 877 (2007)
IV Effect ofSection § 162(m) Under Each Model
A. Arm's Length Model
Under this model, the amount and mix ofcompensation would
generally"be optimal without any intervention by policymakers. In
the absence of§ 162(m), the parties would take into account the
competing issues of(i) shirking and risk-based agency costs and
(ii) manager risk-averseness, andthey would arrive at the optimal
(i.e., shareholder-wealth maximizing)compensation arrangement. The
agency costs push in favor of performance-based compensation, which
shifts additional firm specific risk to the manager.Because of the
manager's risk-averseness, the manager will demand a highermarket
value of performance-based compensation vis-a-vis cash." Under
thearm's length model, these factors would ultimately result in an
equilibrium,with some mix of incentive (risky) compensation and
cash compensation tomaximize firm value.
Section 162(m), which presumably was based on Congress's
perceptionthat the arm's length model is not accurate, upsets this
balance. The provisionpushes in favor of riskier incentive
compensation by making salary (and othernonqualified compensation)
more expensive once the $1,000,000 threshold isreached." At the
margin, this creates a wedge between the type ofcompensation that
executives prefer to receive and the type of compensationthat the
firm is willing to pay. How the burden of this wedge is borne is
afunction of the parties' bargain. If the result is that the
executive simplyaccepts the riskier compensation without any
adjustment in the market value ofthe compensation, then the
executive has borne it by taking on more risk thanhe or she would
like without any compensation. If the result is that the firmpays
the executive a higher market value ofriskier compensation to
compensatethe executive in full for taking on the extra undesired
risk, then the firm hasborne the burden of the § 162(m) wedge.
Finally, the parties could share the
71. As described below, under the implicit contracting costs
theory, the arm's lengthmodel could result in underpayment of
management. This theory suggests that intense publicscrutiny of
management compensation precludes public companies from designing
efficientcompensation arrangements. See infra notes 116-18 and
accompanying text (describing theimplicit contracting costs
theory).
72. Due to significant legal constraints on potential executive
hedging strategies, thisArticle assumes that they will not be
implemented. See supra note 52 and accompanying text.
73. The amount of the increase in the cost of nonqualified
compensation above$1,000,000 is a function ofthe corporation's
marginal tax rate. 1ft is the marginal tax rate of thecompany,
denying the deduction will increase the after-tax cost of
nonqualified compensation inexcess of $1,000,000 (the excess
compensation) by an amount equal to the product of (i) theamount of
the excess compensation and (ii) the ratio of tl(l - t).
-
CONTROLLING EXECUTIVE COMPENSATION 893
burden of § 162(m). In that case, the executive would receive a
higher marketvalue of compensation than he or she would have
received in the absence of§ 162(m), but the higher value is
insufficient to offset fully the burden oftakingon the additional
risk. The effects of § 162(m) under the arm's length modelare
considered in more detail below.
1. Avoidance
If it were costless to avoid § 162(m) by, for example,
disguising cashcompensation as performance-based, we would expect
arm's length parties todo exactly that. This would allow the
parties to avoid the § 162(m) wedge andto retain, in substance, an
optimal compensation design without losingdeductions under §
162(m).
In fact, there are two possible circumvention strategies, though
they arenot entirely costless. First, companies may require the
executive to deferreceipt of any nonperformance-based compensation
in excess of $1,000,000until she retires from the company." Because
§ 162(m) only applies tocompensation received by current
employees," the entire amount paid wouldbe deductible by the
employer at the time of payment. The costs ofimplementing this
deferral strategy stem primarily from the fact that, becausethe
excess compensation is retained by the company (instead ofbeing
paid outto the executive currently), the executive places the
excess compensation at therisk of the company's insolvency or
bankruptcy." In addition, the executiveloses liquidity because she
currently cannot spend the excess compensation, but
74. Upon retirement, the company will pay the executive the
amount deferred plus aspecified investment return thereon. The
investment return is necessary to account for the time-value of
money. Ethan Yale & Gregg D. Polsky, Reforming the Taxation of
DeferredCompensation, 85 N.C. L. REv. 571,576 (2007).
75. I.R.C. § 162(m)(3) (2000) (stating that covered employees
include only those who areemployees ofthe company as ofthe close
ofthe taxable year). A similar deferral strategy wouldbe for the
executive to defer receipt of the compensation until the time when
the coveredemployee's nonqualified compensation falls below the
$1,000,000 threshold. Recently enacted§ 409A would now impose
potentially significant tax costs on this strategy. See I.R.C.§
409A(a)(I) & (2) (requiring immediate income realization for
deferred compensationarrangement that allow for contingent
distributions that are not triggered by the employee'sseparation
from service, disability, death, change in the ownership of the
employer, or theoccurrence of an unforeseeable emergency).
76. In order for the deferral to be effective for federal income
tax purposes, the amountsdeferred must remain subject to the claims
of the employer's creditors. See Treas. Reg. §1.83-3(e) (1996);
Gregg D. Polsky & Brant 1.Hellwig, Taxing the Promise to Pay,
89 MINN. L. REv.1092, 1139 (2005). Had the executive been paid the
cash currently, the executive'scompensation would no longer be
subject to the risk of employer bankruptcy or insolvency.
-
894 64 WASH. & LEE L. REV 877 (2007)
this is not likely a significant cost because one would expect
that executiveswhose pay is subject to § 162(m) (i.e., senior
executives with annualnonqualified compensation in excess of
$1,000,000) would ordinarily besufficiently wealthy and/or
credit-worthy to mitigate substantially any liquidityproblems."
Second, the parties could disguise fixed compensation as
performance-based by using easy to reach objective bonus targets.
The regulationspromulgated under § 162(m) explain that a
"performance goal need not ... bebased on an increase or positive
result under a business criterion and couldinclude, for example,
maintaining the status quo or limiting economic losses.,,78However,
at the time that the bonus arrangement is structured, the
outcomemust be "substantially uncertain. ,,79 Two examples in the
regulations apply thissubstantially uncertain condition. In one
example a bonus based on apercentage ofprofits is sufficient, while
in the other example a bonus based ona percentage ofsales is not.
80 The regulations explain this result by noting that,while a
company is "virtually certain" to have some sales during the year,"
"itis substantially uncertain whether a company will have profits
for a specifiedfuture period even if the company has a history of
profitability. ,,82 In anotherexample, a bonus payable on the
favorable settlement of litigation fails the"substantially
uncertain" factor where the bonus arrangement was establishedafter
the other side has already "informally indicated ... [its]
willingness tosettle the litigation [favorably]. ,,83 These highly
stylized examples are not veryhelpful in giving taxpayers guidance
regarding the application of thesubstantially uncertain
condition.
77. While the corporation's deduction is deferred, this will not
increase the corporation'safter-tax compensation costs provided
that the executive's deferred compensation account growsby an
after-tax rate of return. See Yale & Polsky, supra note 74, at
625. Accordingly, puttingaside the risk of firm insolvency and the
burden of executive illiquidity, a properly structureddeferral
arrangement can avoid the impact of § 162(m), while remaining
economicallyconsistent with the original arrangement (i.e., current
cash compensation) of the parties.
78. Treas. Reg. § 1.162-27(e)(2)(i) (1996). Note that this rule
is conceptually inconsistentwith the regulations' treatment of
stock options, which must have an exercise price equal to orabove
the fair market value of the underlying stock at the time of grant
in order to qualify asperformance-based. Treas. Reg. § 1.162-27(
e)(2)(vi)(A). For further discussion of thetreatment of stock
options under § 162(m), see infra notes 89-90 and accompanying
text, andPart V.A.
79. Treas. Reg. § 1.162-27(e)(2)(i).
80. Treas. Reg. § 1.162-27(e)(2)(vii), ex. 2 & 3.81. Treas.
Reg. § 1.162-27(e)(2), ex. 2.82. Treas. Reg. § 1.162-27(e)(2), ex.
3.83. Treas. Reg. § 1.162-27(e)(2), ex. 4.
-
CONTROLLING EXECUTIVE COMPENSATION 895
Because of the vagueness of this condition, the easy-to-reach
bonuscircumvention strategy creates the risk that the IRS and the
company willdisagree as to how it will be construed, which could
result in litigation andincreased tax liability." As the likelihood
of the bonus payment approachescertainty, this litigation risk
increases. On the other hand, as the likelihood ofpayment becomes
less certain, more risk is imposed on the risk-averseexecutive,
undermining the effectiveness of this avoidance strategy.
Anothersignificant problem with the strategy is that the material
terms of bonusarrangements, including performance goals, must be
disclosed to and approvedby shareholders in order to qualify under
§ 162(m).85 Even if it makes sense forthe firm to set bonus targets
extremely low so as to minimize its after-taxcompensation costs, as
a public relations matter it may be difficult for the firmto
justify bonus standards that are obviously easy to reach."
84. These risks are exacerbated by administrative law doctrine
regarding the appropriatelevel ofjudicial deference afforded an
agency's interpretation of its own regulations. Under theso-called
Seminole Rock doctrine, an agency's interpretation of its own
regulations willordinarily be upheld unless it is "plainly
erroneous or inconsistent with the regulation." Bowlesv. Seminole
Rock & Sand Co., 325 U.S. 410,414 (1945). See also Auer v.
Robbins, 519 U.S.452, 461 (1997) (upholding interpretation and
citing Seminole Rock). Because of thisdeference, a taxpayer could
expect that any reasonable interpretation of the
"substantiallyuncertain" factor would be upheld.
85. See I.R.C. § 162(m)(4)(C)(ii) (2000). Treasury regulations
provide "[t]he materialterms include the employees eligible to
receive compensation; a description of the businesscriteria on
which the performance goal is based; and either the maximum amount
ofcompensation that could be paid to any employee or the formula
used to calculate the amount ofcompensation to be paid to the
employee ... if the performance goal is attained ...." Treas.Reg. §
1.162-27(e)(4)(i) (1996). The regulations go on to provide,
however, that(i) "[d]isclosure ofthe business criteria on which the
performance goal is based need not includethe specific targets that
must be satisfied under the performance goal," Treas. Reg. §
1.162-27(e)(4)(iii)(A), and (ii) a material term need not be
disclosed if "the compensation committeedetermines that the
information is confidential commercial or business information,
thedisclosure of which would have an adverse effect on the
[company]," Treas. Reg. § 1.162-27(e)(4)(iii)(B). It has been
argued that the current § 162(m) disclosure standards
areineffective in ensuring that shareholders have the requisite
information necessary to evaluatebonus arrangements. See
Effectiveness of§ 162(m) in Controlling Executive Pay:
HearingBefore the S.Finance Comm., 109th Congo D914 (2006)
[hereinafter Effectiveness] (testimonyof Steven Balsam, Professor
of Accounting, Temple University) (testifying that
currentdisclosures "lack specificity with regard to actual plan
parameters, targets, thresholds, etc." andsuggesting that
disclosure of this information would "increase the link between pay
andperformance as directors and executives would be less likely to
set low standards").
86. As discussed above, it is not entirely clear that the
current disclosure standards in theTreasury regulations would make
extremely easy-to-reach bonus arrangements obvious toshareholders.
See supra note 85. To the extent that the laxity of an
easy-to-reach bonusarrangement can be hidden from shareholders, the
only constraint on using this circumventionstrategy is audit
risk.
-
896 64 WASH & LEE L. REV 877 (2007)
The most feasible circumvention strategy, therefore, appears to
be the oneof mandatory deferral. To the extent that § 162(m) is
circumvented in thismanner, the provision does nothing except to
impose additional transactioncosts on the parties because ofthe
insolvency risk and illiquidity burden placedupon the executive. To
what extent these burdens are shifted to the company inthe form of
higher compensation is a function of the relative elasticity of
thesupply of, and demand for, individuals capable ofmanaging public
companies.Research using the arm's length model suggests that at a
minimum, a portion ofthese burdens would be shifted to the
company."
2. Substitution Effect
a. Substituting Incentive Compensation for Cash
If avoidance is costly, we might expect to see a substitution
effect wherebythe parties substitute deductible incentive
compensation for nondeductible cashcompensation. In general, stock
options or formulaic bonus arrangementscould be substituted for
cash compensation. There are two advantages ofusingstock options as
compared to formulaic bonuses. First, implementation costsare
reduced because there is no need to negotiate any complex formulas,
toexplain these formulas to executives, board members, or
investors, or to makecalculations to determine payouts. Second,
there is likely less ofan incentive tomanipulate accounting results
because options-unlike formulaic bonuses-arenot based on specific
accounting results (e.g., return on equity) that relate tospecific
accounting periods (e.g., the current fiscal year). 88
On the other hand, there may be advantages to using formulaic
bonusesinstead of stock options. Stock options that qualify under §
162(m) are of the"one size fits all" variety. In order to qualify
as performance-based under the
87. Cf Halperin et al., supra note 9, at 63 (predicting that, as
a result of the enactment§ 162(m), firm profitability would decline
in part because compensation expense would increaseto offset the
additional risk placed upon the executive when incentive
compensation issubstituted for cash compensation).
88. See Reitenga et al., supra note 63, at 1-23 (finding that
firms with formulaic§ 162(m)-qualified bonus plans engaged in
greater income-smoothing behavior than firmswithout such plans);
see also Kevin 1. Murphy, Politics, Economics, and
ExecutiveCompensation, 63 U. eIN. L. REv. 713,739 (1995)
(predicting that § 162(m) "will encouragecompensation committees to
rely heavily on explicit accounting-based bonuses, which, in
tum,will encourage executives to focus on and manipulate short-term
earnings at the expense oflong-term value creation"). This is not
to say that the issuance of stock options does not createincentives
to manipulate. The point is that formulaic bonuses, because they
are based wholly onspecific accounting results, likely create
stronger incentives.
-
CONTROLLING EXECUTIVE COMPENSATION 897
§ 162(m) regulations, stock options cannot ever have a strike
price below theunderlying stock's fair market value on the date
ofgrant. 89 Indexed options-ones whose strike price fluctuates with
a benchmark such as the Dow JonesIndustrial Average-are therefore
not qualified under § 162(m) because if thebenchmark depreciates in
value, the original strike price would be reduced."Accordingly,
fixed at-the-money options are universally used.
As a result, while options cannot be structured to account for
market-wideor industry-wide impacts, formulaic bonus arrangements
can be designed toaccount for such impacts." Similarly, formulaic
bonuses can be structured soas to be based primarily on factors
that are under the control of the executive.For example, if an
executive has more control over the costs of a business ascompared
to its revenues, bonuses could be based strictly on bringing
downcosts. Finally, formulaic bonuses offer greater flexibility
with respect to theamount ofrisk foisted upon the executive.
Because executives are risk-averse,this gives the parties the
opportunity to reduce the size ofthe burden created by§ 162(m).
If the executive were not risk-averse, substitution into either
options orformulaic bonuses could be accomplished without
increasing the cost of theexecutive's compensation. For example,
assume that in the absence of§ 162(m), the parties would have
agreed to a fixed salary of $1,500,000.92
After the enactment of § 162(m), the parties could agree to a
fixed salary of$1,000,000 and a $1,000,000 performance-based bonus
with a 500/0 likelihoodof payment in full and a 50% likelihood of
no payment whatsoever. If theexecutive were risk neutral, this
arrangement has the same ex ante value to theexecutive as the
all-cash arrangement, and the company would not lose any
89. See Treas. Reg. § 1.162-27(e)(2)(vi)(A) (1996).90. See id.
(providing that, in order for compensation attributable to a stock
option or
stock appreciation right to qualify under § 162(m), it must be
based solely on an increase in thevalue of the stock after the date
of grant or award). See also David M. Schizer, Reducing theTax
Costs of Indexed Options, 96 TAX NOTES 1375, 1377-80 (2002)
(suggesting that anargument can be made that indexed options
qualify under the regulations, but ultimatelyconcluding that tax
counsel would not give a favorable opinion on the issue without the
IRSissuing a private letter ruling or amending the regulation).
91. For example, bonuses could be based on the performance ofthe
firm relative to othersimilar firms.
92. Under the arm's length model, it is assumed that this
$1,500,000 fixed salary structureis optimal. Thus, it is assumed
either that (i) there is no need for any performance-basedcomponent
(because, for example, the risk of shirking is extremely low) or
(ii) the marginal costofusing this riskier component outweighs its
incremental benefit. Though this assumption maybe unrealistic in
certain circumstances, it keeps the example simple and does not
change theultimate conclusion that substitution under the arm's
length model invariably imposes anadditional cost on the
parties.
-
898 64 WASH. & LEE L. REV. 877 (2007)
deductions under § 162(m).93 However, as previously discussed,
corporateexecutives are risk-averse." As a result, substitution
will result in an extra costto the parties and the allocation of
that cost between the two parties will be afunction of their
bargain.
For example, assume that the executive, due to her
risk-averseness, wouldbe indifferent between choosing $500,000 of
risk-free compensation and$1,500,000 of potential bonuses which she
has a fifty percent chance ofearning." If, because of§ 162(m), the
parties agree to a $1,000,000 salary plus$1,500,000 bonus structure
(rather than $1,500,000 of salary), then theexecutive is made whole
because she is fully grossed up to account for theadditional risk,
and the corporation bears the entire brunt of the § 162(m)wedge."
If, instead, the parties agree to a $1,000,000 salary plus
$1,000,000bonus structure, then the executive bears the entire
burden of the § 162(m)wedge because she has received no gross up
for the additional risk. If theparties agree to a bonus amount
between $1,000,000 and $1,500,000, then theparties will have shared
the burden of the § 162(m) wedge through a partialgross up.
While the statutory (i.e., nominal) incidence of § 162(m) falls
on the firmbecause the provision denies the deduction for "excess"
compensation-increasing the firm's labor costs-the economic
incidence depends on therelative elasticities of the supply of and
the demand for individuals capable ofleading public companies. The
literature suggests that, under an arm's lengthmodel, at least some
of the burden of the § 162(m) wedge will be borne by thefirm.97 If
so, the result is that, under the arm's length model, to the
extent§ 162(m) results in substitution, the provision decreases
firm value and
93. The two arrangements have the same ex ante market value
because the expected valueof the bonus is $500,000, the same amount
as the excess salary in the all-salary arrangement.
94. See supra note 59 and accompanying text.
95. The bonus portion has an expected value of $750,000.
Accordingly, a risk-neutralindividual would choose the bonus so
long as the cash compensation option was less than$750,000.
96. Note, however, that the corporation is still better off than
had it paid $1,500,000 ofsalary because the after-tax cost of the
extra nondeductible $500,000 salary is higher than theafter-tax
cost of the deductible incentive compensation with an expected cost
of $750,000,assuming that the corporation's marginal tax rate is in
excess of 33.3%.
97. See Halperin et al., supra note 9, at 52-65 (predicting
that, as a result of § 162(m),firm profitability would decline in
part because compensation expense would increase to offsetthe
additional risk placed upon the executive when incentive
compensation is substituted forcash compensation); Murphy, supra
note 88, at 739 ("Executives shifting from salaries
toperformance-based compensation will demand a premium for bearing
more risk, resulting inhigher pay levels. ").
-
CONTROLLING EXECUTIVE COMPENSATION 899
increases executive wealth." As discussed below, these perverse
effects arelikely even more pronounced under the managerial power
model."
b. Substituting Formulaic Bonuses for Discretionary Bonuses
Because bonuses based on objective formulas are preferred under§
162(m), one would expect that companies would consider
transformingdiscretionary bonus plans into formulaic ones. The cost
of this substitution isthat the board of directors loses the
ability to make qualitative judgments (i.e.,judgments not based
exclusively on quantitative results, such as accountingresults)
about the firm's performance in awarding bonuses. In addition,
boardslose the flexibility to alter bonus amounts to take into
account unusualcircumstances, such as war, recession, collapse ofa
major competitor, etc., thatare beyond the control of the
executive.
The loss of this discretion and flexibility must be weighed
against thecosts of forfeiting deductions for the bonuses paid. The
most obvious cost isthe extra tax owed by the company as a result
of the lost deductions foramounts paid pursuant to the bonus plan.
The firm may incur political costs aswell. SEC rules require firms
to disclose in their proxy statements the impactof § 162(m) on
their compensation policies.l'" As a result, firms that
forfeitdeductions must effectively admit to the public that they
are more willing toforfeit valuable tax deductions than they are to
restructure executivecompensation plans.'?' The public may not
appreciate the subtle possibilitythat increased board discretion
and flexibility might be more valuable than thecost of the lost
deductions.
98. Under the arm's length model, assuming at least part ofthe §
162(m) burden is placedon the company through increased executive
compensation, firm value is decreased because ofthe extra
compensation costs. Further, executive wealth is increased because
of the riskpremium they receive.
99. See infra Part IV.B (discussing the likely effects under the
managerial power model).
100. Executive Compensation Disclosure; Securityholder Lists and
Mailing Requests,Exchange Act Release No. 7032, 55 SEC Docket 1352
(Nov. 22, 1993) (requiring that a proxystatement "should in its
discussion of executive compensation policies address the
registrant'spolicy with respect to qualifying compensation paid to
its executive officers for deductibilityunder § 162(m)").
101. Despite the SEC's rule requiring disclosure ofthe company's
§ 162(m) policy, StevenBalsam has noted that disclosures are
sometimes "exceedingly vague" and has suggested thatfirms be
required to disclose the amount of actual deductions forfeited and
the additional taxespaid by virtue of § 162(m) as a result of the
company's compensation practices. SeeEffectiveness, supra note 85,
at 3. To the extent that disclosures are ineffective in conveying
thetrue additional cost of excess nonqualified compensation, the
political costs of forfeitingdeductions would be reduced.
-
900 64 WASH. & LEE L. REV 877 (2007)
If the costs of forfeiting deductions are high, there are two
ways that aboard might attempt to ameliorate the loss of
flexibility and discretion whilestill qualifying bonus arrangements
under § 162(m). First, the board could useobjective formulas that
err on the high side and then preserve negativediscretion to scale
the bonuses back down. One problem with this approach isthat the
objective formulas could be perceived by the public as overly
generousdespite the fact that the board has the power to reduce the
amounts derivedfrom them. 102 Second, the board could narrow the
range ofpayouts generatedunder an objective plan. This would reduce
the impact ofunexpected positivefactors (e.g., sudden supply
shortage for the goods sold by the company) thatare beyond the
control ofthe executive. Consider, for example, a bonus basedon
revenues. Ifthe range ofrevenue growth that is considered in
computing thebonus is smaller (e.g., 80% to 120% of the target),
the effect of unexpectedpositive factors will be muted compared to
arrangements that use a larger range(e.g., 80% to 200% oftarget).
103 In fact, empirical data suggest that companiesthat qualify
their bonus arrangements use a narrower range of payouts ascompared
to companies that do not qualify their bonus plans.l'"
3. Deduction Forfeiture
Because circumvention and substitution may be costly,
corporations incertain circumstances may simply choose to not
comply with § 162(m) andthereby forfeit deductions for compensation
in excess of $1,000,000. Forexample, consider the example described
above where parties are consideringwhether to restructure a
$1,500,000 all cash arrangement into a $1,000,000cash plus
$1,500,000 bonus (50% likelihood of payment) arrangement.
Theexecutive demands the extra $250,000 of expected bonus value so
as to offset
102. Recall that, to qualify compensation arrangements under §
162(m), the material termsof arrangements must be disclosed to (and
approved by) shareholders. Supra notes 37-39 andaccompanying text.
The Treasury regulations interpreting the disclosure requirement
have beencriticized for not requiring the disclosure of
sufficiently detailed information. See supra note85. To the extent
this criticism is valid, the risk of public outrage may not be a
significantconstraint on using this approach.
103. The lower end ofthe range usually will not fall below 800/0
because that is the lowestperformance level that boards feel can
justify a bonus from a public relations perspective. SeeReitenga et
aI., supra note 63, at 2 n.2 (discussing the 80/120
convention).
104. Reitenga et al. found that the rate of earnings smoothing
is higher with respect tocompanies that qualify their plans under §
162(m). Id. at 2. They argue that this is consistentwith narrower
ranges of bonuses, which "result in a greater penalty for extreme
negativeperformance and a smaller reward for extreme positive
performance." Id. This increases theincentive to smooth earnings,
which is consistent with their empirical findings.
-
CONTROLLING EXECUTIVE COMPENSATION 901
the additional risk to her under the restructured arrangement.
In such ascenario, ifthe firm's marginal tax rate is 300/0, the
firm should choose to forfeitthe deduction by paying $1,500,000
cash. 105 If so, the after-tax cost to the firmof the extra
$500,000 cash payment is equal to the pre-tax cost of
$500,000because the compensation deduction is denied.l'" If,
alternatively, the firmpays $1,000,000 in cash and a $1,500,000
bonus (with a 50% likelihood ofpaymentj.l'" the expected after-tax
cost of the bonus is $525,000 (expectedvalue of $750,000 reduced by
the 30% tax deduction). Under thesecircumstances, forfeiting the
deduction makes sense because the tax benefit ofthe deduction is
more than offset by the additional payment that the
executiverequires because of her risk-aversion.
Nevertheless, there may be political reasons why a firm might
choose notto forfeit deductions even in instances when it would
make financial sense to doso. As noted above, the impact of §
162(m) on executive compensationpolicies must be disclosed.i'" As a
result, the public might perceive deductionforfeiture as suggesting
that boards are captured by management. By allowingthe firm to
forfeit deductions, the board implies that it is more interested
inappeasing management (by not forcing them to accept pay that is
more sensitiveto performance) than in preserving valuable tax
deductions. The public likelywill not appreciate that deduction
forfeiture might in fact be the best way forthe company to minimize
its after-tax compensation costs.
A recent empirical study suggests that firms are behaving as
expectedunder the arm's length model.l'" Firms with a riskier
business environment
105. This analysis assumes that the entire economic incidence
falls on the firm (Le., that theexecutive receives a full gross up
for risk). It also assumes that there are no political
costsassociated with forfeiting deductions. See infra note 108 and
accompanying text (discussing thepossibility of political
costs).
106. The after-tax cost of the first $1,000,000 is $700,000
because it reduces corporatetaxable income by $1,000,000;
accordingly, the payment reduces corporate tax liability by$300,000
($1,000,000 x 30%).
107. Recall that the executive would require this arrangement in
order to offset the additionrisk to which she is subject as a
result of the restructuring.
108. Executive Compensation Disclosure; Securityholder Lists and
Mailing Requests,Exchange Act Release No. 7032, 55 SEC Docket 1352
(Nov. 22, 1993) (requiring that a proxystatement "should in its
discussion of executive compensation policies address the
registrant'spolicy with respect to qualifying compensation paid to
its executive officers for deductibilityunder § 162(m)"). As noted
above, however, it has been argued that these disclosures
aresometimes vague and of little help in determining the amount of
deductions lost and theresulting increased tax liability. See supra
note 101. To the extent disclosures are notcompletely effective,
the political cost of forfeiting deductions would be reduced.
109. Steven Balsam & Qin 1.Yin, Explaining Firm Willingness
to Forfeit Tax DeductionsUnder Internal Revenue Code § J62(m): The
Million Dollar Cap, 24 1.ACCT. & PUB. POL'y300, 322-23
(2005).
-
902 64 WASH. & LEE L. REV 877 (2007)
were more likely to forfeit deductions, presumably because
substitution wasmore costly.l'" When the tax benefit of deductions
is higher and when thefirm's compensation practices are subject to
greater public scrutiny, firms aremore likely to preserve their
deductions. 111
4. Director Independence and Shareholder Approval
Requirements
Section 162(m) appears to create incentives for strengthening
corporategovernance procedures because, in order to qualify as
performance-based under§ 162(m), compensation must be approved both
by an independentcompensation committee and by a shareholder vote.
112 Under an arm's lengthmodel, however, there would be no need for
enhanced corporate governanceprocedures because the market would
balance the benefits and burdens oftheseprocedures, ultimately
resulting in the procedure that maximizes firm value.States would
compete for corporate charters by creating default governancerules
that are optimal. If the default rules happened to be suboptimal in
aparticular context, then firms would opt out and voluntarily
require enhancedprocedures in order to best attract capital.
In fact, states have taken a hands-offposture when it comes to
proceduresrelating to the setting and structuring of executive
compensation. 113Furthermore, companies themselves generally have
not opted out of theselaissez-faire default rules.'!" Under the
arm's length model, therefore, thedirector independence and
shareholder vote requirement is unnecessary, servingonly to impose
additional costs on companies who choose to qualify their
plans.
5. Effect on Nonaffected Firms
Under the arm's length model, firms whose base
compensationarrangements for covered executives are set well below
$1,000,000 should be
110. Id.111. Id.112. See I.R.C. § 162(m)(4)(C) (2000) (setting
forth approval requirements).113. See Bainbridge, supra note 50, at
1650-52 (discussing the "rational apathy" by the
states).
114. The exceptions are the recently enacted New York Stock
Exchange and NASDAQrules regulating the corporate procedures for
setting executive compensation, which companieson the respective
exchanges are required to adopt. NYSE, LISTED COMPANY MANUAL§
303A.05(a) (2004); Self-Regulatory Organizations, Exchange Act
Release No. 48,745, 81SEC Docket 1586 (Nov. 4, 2003).
-
CONTROLLING EXECUTIVE COMPENSATION 903
unaffected by § 162(m). Because § 162(m) does not apply to them,
these firmsdo not have to use circumvention or substitution
techniques to preservedeductions. However, empirical data generated
by Harris and Livingstonesuggests that § 162(m) has in fact caused
these firms' compensation costs toincrease,115 which seems to be
consistent with the result that would be expectedunder the
managerial power model. Because Congress has arguably set
abenchmark of $1,000,000 for base compensation, § 162(m) may be
used bycomplicit directors to deflect public outrage over salary
arrangements at orbelow that amount.
Interestingly, Harris and Livingstone view their results from
the oppositeperspective. Harris and Livingstone view negative
publicity regardingexecutive compensation as an "implicit
contracting cost," defined as a "non-contractual cost that third
parties [here, the public,] impose on firms by reactingadversely to
firms' behaviors." 116 The critical assumption made by Harris
andLivingstone is that paying additional compensation always
generates firmrevenues in excess of the cost of the additional
compensation, an assumptionthat would be inconsistent with either a
pure arm's length view or themanagerial power view. 1I? Harris and
Livingstone thus assume that publicity isan obstacle that
undermines the ability of firms to structure executivecompensation
arrangements that maximize firm value. Harris andLivingstone's
results (i.e., that unaffected firms increased their
executivecompensation as a result of § 162(m)) would be consistent
with results thatwould be expected under the arm's length view in a
world where publicity is anobstacle to shareholder wealth
maximization. 118
6. Conclusion Under the Arm's Length Model
Under the arm's length model, § 162(m) simply imposes an
additional coston the parties in compensating executives because,
in the absence of§ 162(m),
115. David G. Harris & Jane R. Livingstone, Federal Tax
Legislation as an ImplicitContracting Cost Benchmark: The
Definition ofExcessive Executive Compensation, 77 ACCT.REv. 997,
1014 (2002).
116. Id. at 998.117. Under a pure arm's length view, the firm
would pay additional compensation to the
point where the additional revenues equaled the amount
ofadditional compensation. Under themanagerial power model, the
firm would pay additional compensation until the risk
ofnegativepublicity was too great, even if the cost of additional
compensation was greater than therevenues generated by it.
118. In contrast, Bebchuk and Fried's managerial power model
predicts that publicity ofcompensation practices increases
shareholder wealth. See generally BEBCHUCK & FRlED, supranote
6.
-
904 64 WASH. & LEE L. REV. 877 (2007)
the parties would have negotiated the optimal compensation
arrangement. Withrespect to this deadweight loss, two issues arise.
First, what is the size of theloss? Second, how is its burden
shared between the firm and the executive?
The size of the deadweight loss depends on how the parties
respond to§ 162(m). The parties could avoid the provision by
providing for mandatorydeferral or easy-to-reach bonus structures.
Alternatively, the parties couldsubstitute qualified compensation
components (e.g., options, formulaicbonuses) for nonqualified
compensation components (e.g., salaries,discretionary bonuses).
Finally, the parties could simply continue their
existingcompensation arrangements, thereby causing the company to
lose deductionsfor compensation in excess of$l,OOO,OOO.
Each of these strategies comes with associated costs, resulting
in thedeadweight loss. For example, mandatory deferral imposes tIrm
insolvencyrisk and illiquidity costs on the executive, while
substitution inefficientlyimposes greater risk on the executive,
the more risk-averse party. The extent ofthese costs is dependent
on a variety offactors.l'" Under the arm's length view,the parties
would choose the strategy that imposes the smallest cost, after
takinginto account the particular circumstances.
Though the nominal burden of § 162(m)'s deadweight loss might
beplaced on one party, part or all ofthe economic incidence may in
fact be borneby the other. For example, ifthe strategy ofmandatory
deferral is utilized, thenominal burden is on the executive because
she bears additional firminsolvency risk and the burden of
illiquidity. However, the parties could shiftthat burden to the
company by simply paying the executive a greater amount tooffset
these costs. Likewise, though the burden of deduction forfeiture
isnominally on the company because its tax burden is higher, the
company couldshift the burden (in whole or in part) to the
executive by paying her a loweramount than it would have paid in
the absence of § 162(m). The economicincidence of § 162(m),
therefore, depends on the relative bargaining power ofthe company
and the executive.
In summary, under the arm's length view, § 162(m) imposes a cost
that isborne by the parties. Unless all of the cost is shifted to
the executive (either inthe form of lower compensation or increased
risk-bearing by the executivewithout any compensation), which the
literature suggests is an unlikelyscenario.i'" § 162(m) reduces
shareholder wealth. Furthermore, to the extentincentive
compensation is substituted for cash compensation, the wealth
of
119. The factors include the executive's liquidity and level of
risk-aversion and the firm'ssolvency, level of business risk,
marginal tax rate, and level of public scrutiny.
120. See supra note 97-99 and accompanying text.
-
CONTROLLING EXECUTIVE COMPENSATION 905
executives is increased because ofthe risk premium they receive.
Both resultsunder the arm's length view are inconsistent with the
intent of Congress toincrease shareholder value.v"
B. Managerial Power Model
Adherents of the managerial power model argue that directors
andmanagement collude to reduce the political costs of
overpayingmanagement. Under this view, would § 162(m) improve
expectedoutcomes?
It is important to note at the outset that the statutory inc
idence of the§ 162(m) penalty is placed on the company because,
without anyrestructuring of compensation arrangements, the
company's tax liabilitywould be increased.V' Therefore, to shift
any part of the economicincidence to the executive, collective
action must be taken by parties. Forexample, cash compensation must
be deferred or substituted into incentivecompensation, bonus
arrangements must be restructured, or compensationlevels must
simply be reduced; otherwise, the company loses deductionsand pays
more tax. As an initial matter, adherents of the managerial
powermodel would be skeptical that actions taken to preserve
deductions wouldeffectively shift the incidence of the burden
created by § 162(m). Theywould likely be even more skeptical that
any actions taken would actuallymake the shareholders better off
than they would have been in the absenceof § 162(m).
This skepticism results from two intuitions. First, recall that
under themanagerial power model, the only meaningful constraint on
executivecompensation levels is the risk of public outrage. In the
absence of anychanged circumstances, therefore, one would expect
executive pay levels
d . hat sti k 123 D . h . han structures to remain somew at stic
y. rastic c anges, WIt out any
121. Put differently, under the arm's length view, there is no
market failure and,accordingly, firms and executives make the
optimum trade-offbetween incentive generation andrisk preferences.
Introducing § 162(m) into the mix upsets this tradeoff to the
detriment of theparties. Unless the entirety of the detriment is
borne by executives, the firm is worse off afterthe enactment of §
162(m) under the arm's length model.
122. See I.R.C. § 162(m)(I) (2000) (denying public companies a
deduction fornonqualified compensation in excess of$I,OOO,OOO paid
to a covered employee).
123. See BEBCHUK & FRIED, supra note 6, at 75 (noting this
stickiness); Iman Anabtawi,Overlooked Alternatives in the Pay
Without Performance Debate 39-43 (Jan. 2005) (UCLASchool of Law
unpublished manuscript, on file with author) (suggesting that path
dependencelimits the flexibility in designing alternative
compensation structures).
-
906 64 WASH. & LEE L. REV 877 (2007)
apparent rationale, would be more likely to trigger public
scrutiny andcriticism. The enactment of § 162(m) provides an excuse
for the colludingparties to engage in significant restructuring
ofcompensation arrangementsall in the name of preserving corporate
deductions. Section 162(m), thus,may create the opportunity for
management to extract even more rents.
Second, § 162(m) provides an excuse for the parties to
substituteincentive compensation for cash. As explained in detail
below, for avariety of reasons, the values of incentive
compensation components arelikely less salient to outsiders than
commensurate levels of cashcompensation.l'" Accordingly, the
managerial power model would predictthat § 162(m) increases
executive compensation levels beyond the amountnecessary simply to
offset the executive's additional exposure to risk,resulting in a
windfall for the executive.
Anecdotal evidence supports these intuitions. It is widely
believedthat § 162(m) contributed significantly to the explosion of
compensatorystock options that began in the late 1990s. 125 Section
162(m) provided anexcuse to drastically change compensation
structures to shift heavilytowards stock options, the value of
which is harder to appreciate than cashcompensation. 126
124. See infra Part IV.B.2.
125. See, e.g., Christopher Cox, Chairman, Sec. & Exch.
Comm'n, Options Backdating:Testimony Before the Senate Committee on
Banking, Housing, and Urban Affairs 2 (Sept. 6,2006),
http://banking.senate.gov/_files/ACFB067.pdf(on file with the
Washington and Lee LawReview) ("[O]ne of the most significant
reasons that non-salary forms of compensation haveballooned since
the early 1990s is the $1 million legislative cap on salaries for
certain top publiccompany executives ...."); James R. Repetti, The
Misuse of Tax Incentives to AlignManagement-Shareholder Interests,
19 CARDOZO L. REv. 697, 709 (1997) ("[A]necdotalevidence suggests,
however, that management is now happily complying [with § 162(m)]
byawarding itselfexcessive amounts of stock options. "); Schizer,
supra note 52, at 468 (suggestingthat "the explosion ofoption
grants is evidence that [§ 162(m)] backfired (or was never
intendedto work)"); Press Release, Fin. Economists Roundtable,
Statement ofthe Financial EconomistsRoundtable on "The Controversy
Over Executive Compensation" (Nov. 24,2003), available
atwww.luc.edu/orgs/finroundtable/statement03.pdf (suggesting that
the increased use of stockoptions during the 1990swas attributable
in part to § 162(m)); Press Release, Senator ChuckGrassley (Sept.
6, 2006) (on file with author) (stating that § 162(m) "seems to
have encouragedthe options culture"). Whether the attribution ofthe
growth in options to § 162(m) is supportedby empirical evidence is
not entirely clear. See Lora Cicconi, Blaming the Tax Code for
theBackdating Scandal, 114 TAX NOTES 1129, 1140 (March 19,2007)
(analyzing the empiricalevidence and concluding that while there is
no question that options growth has beenexponential, "it is not
nearly as clear that the explosion of options comp